STREAM POLLUTION 
371 
From these 5,809 cases, all determinations made at stations where good, mixed, 
fish faunae were found at the time of sampling, were selected for a composite regardless 
of the actual amount of dissolved oxygen present. For purposes of this comparison 
good, mixed, fish faunae were defined as faunae including representatives of the fine 
fish group (trout, or bass, sunfish, perch, and other spiny-rayed fishes), of the rough 
fish group (suckers, buffalo, and catfish), and of the minnow series, in good condition 
Fiqure 6. — The stippled graph is a composite of dissolved oxygen values for all stations at which good mixed fish faunae were 
found, each river system being given equal weight to avoid skewing due to differences in actual numbers of cases. Each solid 
black graph presents all oxygen data for a single river unit regardless of presence or absence of fish at sampling stations. Direct 
evaluation of each river unit can be made in terms of the preferences of good fish faunae from the stippled graph on %vhich the 
black graph is superimposed. 
at the time the. samples were taken. Naturally, the exact species composition of 
these good, mixed faunae varied to some extent with the river system. 
This composite selected as explained, includes 1,297 determinations of dissolved 
oxygen from 372 stations. Figure 6 A gives the distribution of the dissolved oxygen 
values for this composite. To avoid skewing due to uneven number of cases from 
the different localities the various river systems were given equal weights in determin- 
ing the percentages. 
99773 ° — 37 - 
